RESUMEN
Background: The incidence of tuberculosis (TB) in the city of Bandar Lampung, Indonesia, increased during the period 2009–2011, although the cure rate for TB cases treated under the directly observed treatment, short course (DOTS) strategy in the city has been maintained at more than 85%. Cluster analysis is recognized as an interactive tool that can be used to identify the significance of spatially grouping sites of TB incidence. This study aimed to identify space–time clusters of TB during January to July 2012 in Bandar Lampung, and assess whether clustering co-occurred with locations of high population density and poverty. Methods: Medical records were obtained of smear-positive TB patients who were receiving treatment at DOTS facilities, located at 27 primary health centres and one hospital, during the period January to July 2012. Data on home addresses from all cases were geocoded into latitude and longitude coordinates, using global positioning system (GPS) tools. The coordinate data were then analysed using SaTScan. Results: Two significant clusters were identified with P value of 0.05 for the primary cluster and 0.1 for the secondary cluster. Clusters occurred in areas with high population density and a high proportion of poor families and poor housing conditions. The short radius of the clusters also indicated the possibility of local transmission of TB. Conclusions: The incidence of TB in Bandar Lampung was not randomly distributed, but significantly concentrated in two clusters. Identification of clusters of TB, together with its etiological factors such as social determinants, and risk factors, can be used to support TB control programmes, particularly those aiming to reach vulnerable populations, and intensified case-finding.